Columns | April 08, 2010 21:22

Where does the knight go?

Where goes the knight?Australia has a surface of 7,617,930 km², (2,941,299 square miles), making it the 6th biggest country in the world. However, currently there are only three Australian grandmasters on the FIDE rating list (the 4th retired not long ago). How is this possible?

I think I have found the solution to one of the greates enigmas in the world of chess: why Australia has so few grandmasters. And no, I'm not talking about population (around 22 million) or density (which puts the country on a modest 232nd spot). No, I think it's because Australians just don't know the rules well enough. Well, that's one explanation - there might be a deeper one which I'll come to at the end of this post.

It's one specific rule the Australians seem to have problems with: how the knight moves. And indeed, it's not easy. Especially kids, when learning chess for the first time, have trouble with what Donner called "the most paradoxical of all chess pieces". Rooks go straight, bishops along a diagonal, the queen, well, she's very lucky because she can do both, but the knight... How to explain? One straight, one diagonal, that's not easy. My 5 year old niece can't remember it very well, and thinks she can only go forward with her knights (a problem some strong players encounter throughout their career, by the way).

So what's the fuzz about? Well, in the first round of the Sydney International Open, the following happened.

Alan Ansell (1968) - GM Darryl Johansen (2457)
Sydney International Open, 07.04.2010
Position after 31...fxe5

Here Ansell played the illegal (well, to most of us at least) move 32.Nf4-h2?!! and... play continued: 32...Rb4 33.Nf3 (The knight is suddenly perfectly placed for the kingside attack) 33...Nh5 34.Ng5 Nf6 35.Ne6 and the grandmaster resigned.

And this was not just an incident. Last year, in the same tournament, something very similar happened.

Sarah Anton (1718) - Paul Broekhuyse (2118)
Sydney International Open 2009
Sarah Anton - Paul Broekhuyse
Position after 53...Kf6

Here Anton played 54.Nf5-Nd7?!! and Broekhuyse resigned the game as indeed there's nothing to play for when the queen drops.

For other pieces it may be very clear how they move, but many people more educated than my niece have pondered over the question where, and how exactly, the knight goes. In fact, it has been questioned whether the knight actually goes. As Donner argues in Schaakbulletin 60, November 1972, the pieces don't 'go', as 'going' is to follow a path, and along the way all points are passed chronologically. From a geometrical point of view, this is not the case in chess, because 'no points are passed'. So for a bishop on b2 'going' to f6, more than one move is needed. (The bishop 'went' from b2 via d4 to f6.)

More interesting is Donner's point that although optically other pieces may reach much further than the knight, it's in fact the knight that triumphs in a certain way. Whereas queens, bishops and rooks follow lines (horizontally, vertically or diagonally) that are finite, the knight in fact follows an infinite circle! This is what Paul Janse wrote about as well, in a February 1992 article for the club magazine of the Amsterdam chess club MEMO (which was merged into different chess clubs over the years):

The Circle
And Caissa contemplated King, Queen, Rook and Bishop, and she saw that it still was not good enough. She put a cross on the d4 square and asked herself: what are the nearest fields not attacked by any piece I put on the square? She saw that these were the fields b3, b5, c6, e6, f5, f3, e2 and c2. And She consulted the Greek philosopher Pythagoras, who announced her that those squares are at a distance of √5 from square d4. And then Caissa created the Knight, as the piece that precisely covers the squares at a distance of exactly √5. And She saw that these squares describe a circle centered around the knight itself, with the Holy √5 as the radius, and she saw that it was good. The fifth day.

Perhaps it's too easy to explain the happenings in Australia as a lack of understanding of the knight, or even the knowledge of the chess rules. Maybe there's something else going on. Something that would make us realize that the Australian players are in fact much more advanced than we are.

This week Chessbase generously allows chess players the opportunity to solve tough scientific problems, such as the danger of the Large Hadron Collider at CERN generating micro black holes, and the Fermi paradox. In today's post they embbed Richard Feynman's analogy of "understanding nature" and chess, which we'll do here as well, as it might explain all of the above.


Australian chess players seem to have been the first to have realized that in some situations, the knight may go two squares diagonally. Following a fundamental scientific method, they're most interested in the thing that doesn't fit, the part that doesn't go according to what you expected. The Australians seem to have discovered a new revolution in the physics of chess.


Share |
Peter Doggers's picture
Author: Peter Doggers

Founder and editor-in-chief of, Peter is responsible for most of the chess news and tournament reports. Often visiting top events, he also provides photos and videos for the site. He's a 1.e4 player himself, likes Thai food and the Stones.


Peter Doggers's picture

Of course this has happened more often, so clearly the Australian connection wasn't serious, in a not very serious post. It's just that I found it so amazing that exactly the same happened twice in Sydney...

I think timetrouble did the trick.

Levon's picture

But how were these moves actually played and accepted by the opponent?

Clifford's picture

This habit is not confined to Australians. Remember Delemarre's Rb8xc2 in his game against Nijboer in Wijk aan Zee 1995. Again, the GM failed to notice the illegal move. (Pity that Chessbase and Stohl 'cleaned up' the game to remove the illegal move in their database.)
Admittedly it was an Australian who famously resigned after being offered a draw - Sztern-Lundquist 1984.

Jeremy's picture

I am from Australia, and I think the article is written in good humour. I just hope that other Australians don't get offended!

Bert de Bruut's picture

Peter, haven't you got it all wrong? Didn't you know that in Australia they don't have knights but they play chess with kangaroos? Now where is that beer?

Castro's picture

"so amazing that exactly the same happened twice in Sydney"

Exactly the same happened a tousand times in every chess club, Peter. (Unless it was born yesterday).
In competition, its rare, but do happen in every country, I'd say every year. Maybe that and castling illegaly are the most popular mistakes (or cheating). And there are some famous examples!

As for cheating, the one I find most hilarious is fianchetting a bischop in a single move and with a single hand (Say, you grab the bischop in c8 and move it onto b6 while also pushing the pawn b7-b6 with your finger. Everything very quickly, of course.) It's so stupid it becomes funny.

As for Australia, could it be because their most famous "springer" animal isn't the horse? ;-)

Nonationalism's picture

Great piece! Pity some folks on here need to look up the word humour in their dictionary before understanding.

hugo's picture

austrians are great peoples! with Arnold capablanca the great player!

Castro's picture

"name me one FIDE rated tournament where exactly the same illegal manoeuvre appaered two years in a row, so that in fact these game couldn’t be entered correctly into a database"

I couldn't :-)
First, I'm not used to enter moves in databases (maybe I should). Who knows someone used to that could answer you positively.
Second, I was merely making an intuition and statisticaly intuitive exercise there. I witnessed that exact manouver once (or twice?) in official competition. I've witnessed a few other kinds of illegal moves in serious play. I dont have many continuous years of active playing. So, my intuition says it happens once and a while everywhere. Maybe there is a FIDE tournament with that illegal manouver apearing twice in the same... round (I thought "game" was asking too much, but who knows?)?
There is a problem in knowing that, namely that those events must be corrected, and lots of times one just have access to the corrected data...

Rob Brown's picture

In the featured Australian games the horse was not a knight but a kangaroo, which, as devotees of Aussie Chess know, can make a short bishop move once each game!

Rob Brown's picture

One wonders how Anton and Ansell would fare in Seirawan Chess where Hawks fly about the board like knights and bishops combined and elephants stomp about like rooks ridden by knights. The mind boggles.

Peter Doggers's picture

“Exactly the same happened a tousand times in every chess club, Peter. (Unless it was born yesterday)."

Believe me, I've been to chess clubs. I know.

"In competition, its rare, but do happen in every country, I’d say every year."

Of course this is obvious as well. But name me one FIDE rated tournament where exactly the same illegal manoeuvre appaered two years in a row, so that in fact these game couldn't be entered correctly into a database.

Eiae's picture

Russian champs going on...

Clifford's picture

I think there is a fundamental misunderstanding here of how a kangaroo moves.
A better example would be Zhao-Rujevic, Australian Championship 1999, where White played Qh5xb5, jumping over a pawn on g5. Because the game was played in Australia, Black accepted the move and White went on to win the game and finish second in the Australian Championship. White later became a GM (under traditional Northern Hemisphere rules).

antichrist's picture

And it's not just knights they have trouble with - they also can jump over pawns with rooks.

Björn Verstraate's picture

I remember a team match a couple of years back, where one of my teammates decided the match in a very interesting way.

The position was basically a pawn endgame, in which both players had a passed pawn that couldn't be caught by the opposing king. However, both pawns were on the third rank, so they'd promote simultaneously, ending the game in a draw.
Both players were in quite a bit of time trouble, and my friend decided to play the move h3 - h5?!!. The opposing player did not assume anything wrong, and moved his pawn up a rank as well. Of course, the pawn race was now won for our team. The match eventually ended in 4.5 - 3.5 for us, so it was crucial for the end result as well. The look on our player's face, when we explained what he had done, was priceless :-).

The white player was Jeffrey Stoffers, but I don't remember who was playing black. It's not as strange as the GM cases from the article, but I figured I'd share. :-)

Rini Luyks's picture

Maybe an idea for an Australian version of the game (like Rob Brown): transform ONE of the knights into a kangaroo, hopping like in the two examples given: Kf4-h2 or Kf5-d7+...

Michael's picture

Another scientifically interesting game was Agdestein-Shirov, Bundesliga 1999, in which Black suddenly had two dark-squared bishops. Agdestein clearly remembered that one bishop had gone from c3 to d3 (!) and later to d4 (!!), so the bishop on g5 must have been the one that was on d7 earlier. I wonder what Feynman would have said about this case.

jo snow's picture

I read somewhere that Tal when learning Chess played a variation ( the local rules) that Knights could only move forward, which was put forward as one of the reasons his play was so aggressive.

If I remember correctly the article showed a few games where he missed opportunities/ correct defenses because of this.

Its so damn difficult to unlearn chess!!

PS. I always thought Knights jumped - they didnt go

Rob from Sydney's picture

I'm an Australian Chess player and I did find this article amusing. The lack of chess strength has more to do with it's isolation (especially the huge distance between the european chess population) and our climate. People over here spend most of their free time outdoors going to the beach, play outdoor sports , watching sport etc. Australia traditionally do very well in the Olympics (finishing in the top ten often, on medal count) and passing many countries with much larger populations.

As far as GMs are concerned, I believe after this Easter, the total number of GMs will be 5, with young George Xie getting his third and final GM norm. I see Australian Chess getting stronger in the future due to our close proximity to China and India - both emerging chess super powers.

jo snow's picture
Billy's picture

Hello There. I found your blog using msn. This is a very well written article.
I will make sure to bookmark it and come back to read
more of your useful information. Thanks for the post. I will certainly

Atheistic Bishop's picture

You are nationalist, Peter.

Frits Fritschy's picture

Thanks, Peter, now I know why I never became grandmaster.
Some 20 years ago I played first board in a team competition for companies against someone with at least 600 elo points less - if he had an elo ranking at all. He had played well, but after his knight had taken a pawn on a7, I played c7-c6 and the knight was going to be picked up by my king (or something like that).
He found the remarkable solution Na7-b8, and suddenly he had a way out on a6.
My team mates kept me wondering what I could have missed and had a good laugh on their top player after I had lost the game...

Felix's picture

I would expect those things to happen if one player walks away from the board and comes back not remembering the position before. If such an illegal move is executed, I would guess this is much easier to spot it during the execution.

However, if you notice you made an illegal move, don't say anything and wait until you make a mistake. Then go to the arbiter and tell him about your move - it's like an additional life :)

(an easy way to make friends, especially in an open tournament when all rounds will start later therefore :) )

David Smerdon's picture

There was also a famous Australian blitz championship where the title was decided by an obscure king-and-pawn endgame. As each player's rook pawns raced towards promotion, black, a move behind, found an amazing way to win: 1.h4 a5.25!! 2.h5 a4.75!! 3.h6 a2!! 4.h7 a1=Q 5.resigns.

Ansell's Dad's picture

I realise this article is intended to be humorous (and it is I think), however, I would just like to add that I witnessed the resignation (but not the illegal move) in the Ansell-Johansen game. At that time Johansen had over 20 minutes left on his clock and Ansell just under 2 minutes. So Johansen can have been under no real time pressure when the illegal move was made, though Ansell probably was.

Futhermore, in response to some claims made elsewhere that "Johansen lost BECAUSE of an illegal move": Rybka, after 20 minutes evaluation, scores the position -0.17, before the knight was moved and +1.20 immediately afterwards - so the illegal move gifted Ansell a little more than the equivalent of a pawn in an equal position. It was Johansen's reply 32. ... Rb4 (after which Rybka scores the position > +5) that was the real game losing move, and it would seem reasonable to conclude that had Johansen evaluated the position adequately after 32. Nf4-h2, he would have made some other reply (Rybka suggests 32. ... Rf8) - I would imagine that most GM's would opt to play on when down only a pawn, with a healthy time advantage, against 1900/2000 rated opponents? So Johansen not only missed that the move was illegal, he also missed the threat it posed. That is two mistakes.

Finally I would like to say that Ansell was very upset himself, when he learned what had happened, and I believe the rest of his tournament suffered accordingly. Well before round 2 was paired he actually asked the arbiters to have the result in the game overturned, but this was refused. He was told his only recourse was to appeal.

Clifford's picture

I would have thought that the shock of seeing a totally unexpected move - in this case actually an illegal one - would mean that Johansen did lose precisely because of the illegal move. In a better position he played the move which he had prepared and which would have been very effective against the legal Ne2.
Pity that the arbiters did not allow the result to be changed if both players wanted to agree to a draw.

Hortensius's picture

Fantastic Feynman!

santa's picture

I read in one of Barden's books about a player castling with the rook from the next board. The game was over before the owner noticed its loss!

Caitlyn StClair-Trethowen's picture

I am not surprised Miss Anton played her knight like a bishop. She played illegal moves five times in my boyfriends game against her, once moving the bishop like Knight in fact!
I am surprised by GM Johansen not picking up on his opponents move though.

Your comment

By posting a comment you are agreeing to abide our Terms & Conditions